← Embracing Digital Transformation
Loading summary
Transcript66 lines
- [00:00]
A
Welcome to Embracing Digital Transformation. Before we dive in, I wanted to personally thank you for listening. Many of the ideas we discuss on this show inspired my new book, AI Augmented Teams. If you're looking for practical ways to combine human expertise and AI to achieve better outcomes, I think you'll find it valuable. Learn more at Paydar AI Books. That is P A I D A R AI Books. Now let's get started with the show
- [00:30]
B
with the purpose of digital transformation. I mean, why are you doing it? You're doing it because you're hoping that data will help you. You're hoping that data will in the future make the decisions more. You're hoping that data will allow you to automate things. And then it turns out, and I mean, you've also been for so long in digital transformation, right? There's always that big hope, that big promise, and then people look back at that transformation and just say, there's all of this frustration. I was promised transformative results.
- [01:04]
A
Why this matters to you? Most transformation programs still treat data as a reporting layer instead of a decision discipline. And that is why so many initiatives stall after the dashboards go live. If your teams are using metrics to defend positions instead of test them, you do not have a data problem. You have a leadership problem. In this episode, we look at the difference between collecting information and changing how an organization thinks. That distinction affects culture, accountability, and the pace of change. Let's get into the conversation. Welcome to Embracing Digital transformation. This is Dr. Darren on this episode, how to Build a Data Inspired Decision Culture with Sebastian Wernicke.
- [01:53]
C
Sebastian, welcome to the show.
- [01:55]
B
Hi, Darren. Great to be here.
- [01:57]
C
Hey. I'm really excited about this topic because not a lot of people really use data to help with their cultural change. A lot of times they go, let me see which way the wind is blowing and try and steer through this very complex digital transformation, which includes a lot of people, right. And culture change and all that. So I'm excited to talk about this today, but before we get started, everyone that listens to my show knows that I only have superheroes on the show, and every superhero has a background story. So, Sebastian, what is your background story? What's your origin story?
- [02:41]
B
Yeah. No radioactive spider bites or anything like that? No. Darn.
- [02:45]
C
I was hoping for something like that.
- [02:47]
B
Well, maybe one day. Maybe. That was my hope. So when I got started getting into data, I was actually studying bioinformatics because I was always super, super interested in, you know, what can data tell us about the human body and what can it do with helping us Figure out how everything works and how we can help, you know, find help when the body is not working. And so that got me super, super interested in data. But then unfortunately, I'm just an extremely curious person. And at the same time, I had an issue when I graduated from bioinformatics that there was only one data set. It was the time where we had that single human genome, or as it later turned out, that draft sequence. And so that got me curious about all of these other things. So I did data with financial services, with industrial clients. And so for, well, 20 years now, I've basically tried to help clients to do better things with their data and find the value that they're hoping for.
- [03:48]
C
So that's, that's really interesting. I bet you see data in everything that you do then.
- [03:53]
B
Yeah. And sometimes where other people also don't see it. Right. So I will actually look at, you know, let's say there's a field of wheat outside of there. And I've recently worked with somebody in precision agriculture. So for me it will be like almost in my mind. I divided up, oh, it's all these little squares where you could measure the temperature and the humidity. And I recently read about this really, really cool machine which I think is. It sounds like sci fi. So if you're dreaming about a machine that can identify weeds with AI and then use a laser to shoot them down, that exists, it's a million bucks. But you can almost, you know, you can buy one if you want one. And so living in this amazing world where if you just look closely enough, data is everywhere. It's like background music, you don't notice it, but once you start paying attention and basically you say, you know, yeah, what's, what's outside there? And you know, oh, the cars, they're all collecting data even if they're not self driving. And if I go grocery shopping, I mean, how the store arranged it, the pricing, how the products are made, everything is data. Data is already running the world if you just look for it.
- [04:58]
C
Well, there's one area though that we don't use data really effectively. From my perspective, which is our topic today, which is I don't think we use data very effectively in business transformation or in digital transformation because it's messy, especially with people. When you get people involved and you talk about cultural change, things get really, really messy and people just kind of. It's weird because I've been involved in a lot of these major digital transformations and it's like everyone's just going from their gut and I'm like, we gotta have some kind of data to help us, right? But I don't even know where to start.
- [05:42]
B
Well, yeah, and so I think there's many facets to that, to unpack there. Because I think even with the purpose of digital transformation, I mean, why are you doing it? You're doing it because you're hoping that data will help you. You're hoping that data will, in the future make the decisions more. You're hoping that data will allow you to automate things. And then it turns out, and I mean, you've also been for so long in digital transformation, right? There's always that big hope, that big promise, and then people look back at that transformation and just say, there's all of this frustration. I was promised transformative results, and it's not like things aren't changing. You know, you might have that dashboard, you might have Bill. Dashboards, yeah, thousands of dashboards, right. It's almost like people are saying, oh, I want more data. And then they have all these dashboards and they're saying, what am I supposed to do with all this data? So it's clearly not going in the right direction, and it's clearly not applied where it's actually helping. And I think that is sort of the question I got very interested in for a long time now is why do we find it so hard to get transformative change out of data? It seems very easy for us to get optimization out of it. Right? So I can basically say I use data to improve that process by a couple of percentage points. It's easy for us to say, you know, I have this big meeting coming up. Let's find some data that will sort of support my position. But we're not using it really effectively to find that position or to find a novel position. It's rather saying, you know, okay, let's see how we can make this seem a bit more factual than it is. And I've been in that position myself, so not pointing fingers, I've also done that. But I think that's what we need to overcome. Because ultimately, if all data is doing is taking us forward in these incremental little optimization steps, then I think you're just being too slow in evolving as a business these days. Right. I mean, we have all of these supply chains breaking down, complexity happening.
- [08:06]
C
I, I want to, I want to follow on, On. On where, where you were going there, which was. A lot of times we use data to reaffirm our position instead of looking at data to help us figure out a position. Yeah, but I, I know when I was working my PhD dissertation, that's what we were taught to do was come up with our hypothesis and then come up with data to disprove it or to prove it. There, there's, there's a whole bunch of different, you know, things around that. But isn't that innately what we try and do with data? And how do we get out of that mode where we can actually start using data to. What's the right word? To help us make decisions instead of using data to just reaffirm what we've already decided?
- [08:57]
B
Yeah, I think we all intuitively have this thing that I like to call the data deficit theory. And it's this thing that we think if only we had the right data with the right people at the right point in time, decision making would just improve on its own. And I think you just said a very interesting word there. Right. It's how we were taught. I mean, you were taught as a scientist and so was I. So I think there's already a little hint in there that tells us, well, probably that's not the natural way you would use data. I mean, if you just said, here's your hypothesis, what would be a natural inclination? Well, I want to prove that only when you're taught as a scientist or when it comes naturally to you to think like a scientist, are you thinking about falsification and saying what could be wrong here? And it's also actually, if you're doing a PhD, it's part of your incentive. And so I think that there's a couple of points there that already give us a hint what might be going wrong. If an organization just tries to live on that data deficit theory, you need to teach people to think like that. You also need to set the right incentives to think like that. Right. If I'm going into a meeting and there is no good reason for me to disprove my hypothesis because it'll get my project killed, it'll make me look stupid, all kinds of reasons. If I don't have that environment around me, then my natural tendency will always be to say, I'm going to use data that supports me. And add to that, I think, a couple of decades of psychological research that tells us exactly the same. If you confront people with some data that agrees with them and some that disagrees with them, the natural tendency is not to say, oh, let's do a very objective evaluation of the situation, weigh this data carefully against each other. Of course not. Right. You're going to say, well, the data that agrees with me. That's good data. And the other data, kind of sketchy, you know, not sure about the source, the analysis and so on.
- [11:03]
C
So how do I do I have to hire someone externally without the same motivations then, or without the same bias? Because we all have bias based off of. No, I'm kind of serious on this because. Yeah, yeah, well, also to give you some business, right, Sebastian? Right. Well, but I mean, how do we overcome, overcome that, right? Because hey, if I want to make a decision or is that the job of the CEO, I don't know where he needs to sit back or the leader needs to sit back a little bit and say, hey, I got to make better judgments based off the data instead of coming.
- [11:43]
B
I think it's definitely, yes, I think it's definitely a cultural decision that you're making at some point that you basically say, I want my company or I want my department or I want my team to run like that. And we can see this with a couple of companies, I think, that have made very deliberate choices about how they're going to deal with data. So some examples people might have heard of is if you join Amazon, for example, they will give you a lot of training. I mean, some people go as far as calling it indoctrination, right, because it's so rigorous, but they basically tell you, this is how we deal with data, this is how we make decisions. Right? So famously, for example, you have to write these six page memos, very carefully argued with supporting data and also disagreeing data. And I mean, you might agree or disagree with a six page memo per se, but I, I think what this tells us is this company is very deliberate about telling you once you come in what to do with data and what the expectations are, how you bring data to the table and how you discuss it and how you deal with it. And unless you make that conscious decision, which ultimately, I mean, going back to your point, I think is very much a leadership decision because, well, who sets the culture, who sets the cultural tone that is set by leadership, that is, and not by the posters that are on the wall. Right. But, but living by example, the lived experience of people.
- [13:12]
C
So, you know, somebody incentives that they bring to their employees. Right? I mean, yeah, yeah, we all know this. If I'm incentivized to make mistakes, I'm going to make mistakes. If I'm incentivized to cover things up and to only show green on my dashboard, that's what I'm going to show. Right?
- [13:33]
B
Yeah, exactly. And if you think about the dashboard. I think that's a great example. I mean, what would be the two scenarios or sort of the counterfactual here? In one world it's all about the green dashboard, right? And as long as the dashboard is green, everything is good and the yellow blip appears and then the finger pointing starts and it's sort of whose fault is this? And in a company that I like to call a data inspired company because they are sort of willing to deal with data in a more curious way, they would look at that dashboard and say, okay, that's interesting. Let's first find out where that yellow blip is actually coming from. And let's not be satisfied with the very first answer, which is obviously that some number went outside of some boundary that you've given it. But, but let's dig a little deeper. Let's think about root causes here. Let's really think where this came from and understand is there anything we can do about the system and not have this knee jerk reaction of things need to go back up sometimes, of course, I mean I'm generalizing, sometimes that is the right direction. But in general data should make you curious and not always be perceived as a problem to be solved immediately.
- [14:54]
C
Well, this sounds very Edward Dimming like of you Sebastian, because I mean that's how he kind of started the whole Japanese economy. And the Japanese are really good at this, at using data to help optimize process and affect change in factories and things like that.
- [15:16]
B
I mean, big Deming fan here as well.
- [15:18]
C
Oh great.
- [15:20]
B
One of the things I found out about Deming, which I always find very interesting, I mean you might have heard that saying of what gets measured gets managed or what gets measured gets done. And the interesting thing is a lot of people attribute that to Deming which unfortunately I think makes him turn around in his grave because it's one of the, it's one of the clearest mis citations I have ever come across. Because his real quote of course is it's wrong to assume that if you can't measure it, you can't manage it. He'd called it one of the seven deadly sins of management. And that's what I so much appreciate about Deming because, you know, the inventor of statistical process control, you might think he's the numbers guy, he comes in and says you need to measure everything. And we want the, you know, the five nines, the six nines, the seven nines, precision and everything. But no, no, no, no, no, he's, he's very precise about saying this is where you should use numbers to drive things forward. And here are areas I'm very clear about where numbers don't make sense. And we are calling upon the humans, upon the leaders to actually bring their non numerical skills to the table. And I really love that about Deming.
- [16:36]
C
You know, that's really interesting that you said that because that kind of fits in with the whole AI movement that we're seeing as well. Think about that, right? If AI every. Everyone's afraid AI is going to take their job. But remember, AI is just a number cruncher. It's a statistical analysis or statistical program that does probability weights. That's all it does, right? We still need humans. We still need humans to make decisions. And so this all kind of folds in together. What we're seeing happening today with, I've already started seeing a lot of AI projects are failing and I think it's because we've kind of relinquished ourselves to the machine. But we still are involved, we still need information so that we can actually make good decisions.
- [17:29]
B
But you're raising a super interesting point here, which is I think it's again coming down to that clarity of decision making. So where data goes wrong is of course when you say, well, I want to be data driven and you mean that in a very literal sense. And you assume that any decision can be decided by data alone as long as we bring data to the table. Which of course goes against what Deming said where he said, well, there's some data decisions, there's some non data decisions. And I think with AI it's exactly the same that we have to be very deliberate about thinking what are decisions that AI is good at, where can AI help and what are decisions that AI is not good at? And I'm seeing a similar phenomenon like you describe. It's essentially just a fear of missing out and oh, let's get as much AI as we can. And essentially the AI strategy itself is just repeating the word as often as you can without actually saying, what's our real business problem here? What are we trying to do for the customer? What are we trying to address in the market? And then the footnote should be let's use AI where we can to make that happen. But it's flipped on its head right now.
- [18:45]
C
Oh, it, it totally is. And I'm hoping that we get back to some reality. And I think we will, I hope that we will. But it's going to require some real leadership from our companies, both small businesses and big corporations and government and education. This is a big deal. What we're going through right now.
- [19:12]
B
And it's so interesting, I think with the whole AI hype, this is what I like to tell people. Everybody right now is so focused on the supply of AI. All the billions or, well, trillions now,
- [19:23]
C
even trillions of dollars. Yeah.
- [19:25]
B
God, these big numbers being poured into data centers, electricity, land. And so. And all of these things are of course important for the supply side, but I think we need to be thinking a lot more about the demand side. How are corporations going to absorb that? Are they prepared to absorb that? Is their culture prepared to absorb that at the scale that the supply is being created? And once we exit the age of token maxing, where essentially you just said that simple incentive make the CPUs glow red hot, right? Yeah, yeah, that's the incentive. Right now we are going of course to turn around and say, well, where did it actually bring value? And those that have been deliberate about it I think will find that actually you don't need to max the tokens, you just need to use them at the right place in the right way. And we will find, I think we will see a rebalancing there for sure.
- [20:21]
C
Oh, we're already starting to see it. In fact, we're going to go through what the trough of disillusionment that's happening right now. I'm already seeing it. I'm already seeing a pushback and some of the big token bills are coming back. Like, oh my goodness, I spent $5,000 on tokens this month.
- [20:41]
A
Yeah.
- [20:42]
B
Although there's of course a bit of irony. You know, you tell people, use as many tokens as you want and then you're surprised about the bill. I mean, did you think they were free? Well, no, they're not. Right. But it's rushing in is what we're seeing. I think that's what we're describing. Right. Is rushing in. The knee jerk reaction. Let's do the first thing that comes to mind and not thinking it through all the way to the end, like where do we want this to end and what does a transformed company actually look like? And how is that better than where we are today and how do we get there?
- [21:15]
C
Well, I'm glad you brought that up because I'm trying to address that exact problem with my book that just came out called AI Augmented Teams. And I have another one out coming out called AI Augmented Organizations which tries to answer that exact question, like, what do I look like when I'm truly AI augmented, not AI driven humans that are AI augmented, which is very data driven, which we need. Right. How do I know if I'm, if I'm being effective and efficient? Right. With what I'm doing? It all smells a lot like the turn of the millennium when the Internet came out and then five or six years later when cloud computing became really big. It seems like the same cycle.
- [22:17]
B
Oh yeah, it's history repeating. I mean, with AI for me, it very much reminds me of some may still remember the term the big data era.
- [22:27]
A
Right.
- [22:27]
B
That we had 10 or 12 years ago. And then you must have been loving life then.
- [22:32]
C
Right? The big data, that's me.
- [22:35]
B
And now it's big. Yeah. And then, you know, a bit of the analytics and everybody got excited about machine learning. That was a phase. Right. And then all the machine learning departments got renamed into AI departments and so on. And I think it's always the same theme of you rush to get the technology in. And just to be clear, I mean, obviously the technology is important to make it happen, but unless you think about how does that integrate with what I like to call the decision architecture of a company. So where are decisions made and how this isn't going to work? Because in essence, all of these revolutions, if you will, or the promises at least of them, were about saying, we're going to hand over more of the steering wheel to data. Right, Right. We are going to hand the decision making to the data in an ever increasing way where first with maybe business intelligence, we said we're going to use data to find out what's going on and then humans are going to make the decision. Then with machine learning, we went a step further and said, well, probably we're just going to let data decide on its own in these very, very well defined environments and with a very clear goal. And now with AI, I mean, if you look at how some people talk about agents is sort of, well, we'll just let AI figure everything out. We let it figure out what decisions it should be making and then what the decision is at the end. And so if you're not deliberately thinking about, well, how do I want decision making to look like that is going to fail? Because then you're going to bring in the technology, you're going to apply it to whatever's close by or where it seems easy to do and you're not going to get those long term benefits out of the whole thing.
- [24:29]
C
Well, so this, this leads to a question I had around. Because you're a data expert, there are multiple. Right. Decisions that could be made from the same data.
- [24:40]
B
Correct.
- [24:40]
C
There isn't just one path that can be taken.
- [24:44]
B
No, no, absolutely right.
- [24:47]
C
So I think this is myth that
- [24:48]
B
this could be done.
- [24:49]
C
Yeah, yeah, exactly right. So I think a lot of people feel like AI is going to tell me the right decision and I'm like, no, no.
- [24:58]
B
But data is a place of comfort that people like to turn to. I think this is, and this has been an issue with data for quite some time. Right. If you hear a leader saying I want more data, well, in a way they are saying they would like more data. But why are they saying it? Because they feel that the decision they're making is so complex and so high stakes that it's something that makes you feel uncomfortable and that makes you feel insecure. And so you're just hoping to get data to do that for you. And I think it's similar with AI. And the point you're raising for my book, I constructed an example actually of that where I had an example of three TV shows and you're supposed to choose which one is the best one. And it turns out if you're trying to do that in a data driven way, there's a very good, objective, solid argument to be made for each one. The data doesn't decide that. So you might say, I'm going to go for the one with the best average rating that's valid, that's going to give you one of the shows. You might also say, I want to have the show that has the least bad ratings so that are not going to disappoint people. That's another one. And it's a perfectly valid argument. And then there's a third one where it's about sample size. So one of the shows has been rated by only a few people, another one by a lot of people. And you might just very fairly say we're going to discard the show that only has a few ratings. Right. That's not very, very certain yet how that is going to turn out. And so I think, and this is of course just an illustrative example, but it's there to tell you exactly what you're saying. The data itself will not tell you what the outcome is unless we are again, in a scientific setting where we designed the data very specifically to answer what we're looking for. Even then, though, what do we end up? We end up with a P value. Right. We are never 100% certain it's a percentage.
- [26:54]
C
I'm glad you brought that up because that assumes closed systems, very well defined constraints and in business it's not very well defined. Right.
- [27:08]
B
CUSTOMER CHATS for repeatable experiments. Right? You're not going to make that big strategic decision 20 times and see how it turns out. You have the one shot.
- [27:17]
C
So this is why I think I'm drawn to AI Because AI can deal in the fuzzy, but it can give me, if I use it correctly, I can bounce ideas and different paths very quickly. That can be data driven and say, what if I go down this path with this data that I have compared to this one, compared to this one, give me the information back so that I can now make a good decision based off of my bias? Because I think all decisions are made off of bias. I just, I know that's contrary to what I was taught as a scientist.
- [27:58]
B
Oh, no, no, no, no, no, no. This is psychological research Would agree with you. There are very interesting brain studies from the 1990s where they studied patients that didn't have an ability to process emotions. So they had a certain damage to the, an area of the prefrontal cortex. And it turns out these people were, if you were talk to them, perfectly normal, right? So, so smart. You can have a normal conversation. But these people have one thing they cannot do. They cannot decide. So it turns out that even if they are supposed to choose, what am I going to have for lunch, they will not be able to make that decision. They will go back and forth on the advantages, the disadvantages of all the options. So I think emotion is what makes decisions possible. This is how we humans operate. So just the notion that we can have a purely rational, like you say, bias free decision brain research would disagree with us. This is not how we work.
- [29:03]
C
Sebastian, you're destroying the need for data. No, I should just go with my gut. That's what I'm learning.
- [29:10]
B
Yeah, exactly. Yeah, yeah, we said it all along, right? No, of course not. And I think this is where it then comes down to, why is data useful? Because I think you already gave it away. Right? Data is super useful to make informed decision. It can help us play through scenarios, it can help us understand certain aspects of a decision. Right. How big is that market that I'm entering? What are the likelihoods of failure? Are there examples that have worked that have not worked? All of that is very, very useful data and we can use that in the decision making. But what we shouldn't expect from data is to make the decision for us. So I like to always say data makes decision making a lot better, but not easier at all.
- [30:02]
C
Oh, I like that. I like that a lot. Because what that tells me is it requires me as a Human to use the brain. Right. Which I feel a lot of people are kind of. What's the right word? They're afraid that AI is going to take away our ability to think. I actually think it's going to force us to think more because it's going to give us more options than what maybe we had before and expose some of those cases that we can't see ourselves because our bias has limited us. So I think it may actually expand our ability to think beyond, you know, our little box that we sit in all the time. Yeah.
- [30:47]
B
Although, I mean, AI does something very sneaky, I think, which is, you know, it can produce because of that property, you know, that it will always give output. I mean, that's what it's trained to do. Yeah.
- [31:04]
C
Yeah.
- [31:05]
B
And so when I interact with a human and they haven't thought things through, usually I can tell by the sound of it that they haven't thought things through. And of course, with AI, I personally find it so hard to have all of this material being generated. That sounds so good. But you need to think much, much harder than with a human written text about, is this really true? And yes, because humans.
- [31:32]
C
Uncertainty. Right. Humans inject uncertainty in the way that they present something or they write something.
- [31:38]
B
Yeah.
- [31:39]
C
Where A.I. you're right. A.I. is confidently wrong.
- [31:43]
B
It's confidently wrong. And then you almost need that Spidey sense almost of saying, you know, this sounds wrong. Why? Why? Why? And then at some point, you know, just dawns on you and say, oh, this is a category error. It's one that sounds so plausible. But it's just. And AI of course, makes category errors all the time, strings together these things. So it's. I would fully agree that AI makes us use our brains, but at the same time, it's forcing us to do very, very hard work, I have to say, you know, reading through an AI generated text and making sure it's really good, I find that very, very exhausting because I have to be, you know, careful about so many things, and I have much less difference between the signal and the noise than I would have if a human had written that. Right. If it's noisy thinking, it's noisy writing. And now I get noisy with clear writing. Oh, that is. That can be painful.
- [32:38]
C
Right on the nose. I know. It's described it so well. So. Hey, Sebastian, this has been great talking to you, but we're out of time if people want to reach out more and. And you got to pitch your book because your book, your book's fabulous. Right? So tell, tell people a little bit about your book and how they can reach out and. And find out more from you.
- [33:00]
B
Yeah, absolutely. So while I was working all of these years in data, I of course noticed what works and what doesn't. And oftentimes I found people work so much for the data, but the data isn't working for them. So what can be done about this? And all of these insights over the years, they then made it into a book. And the book is called Data Inspired. And the reason it's called that is because I want to get away from that notion of data driven, which, if you take it literally, is. Exactly. I will let the data make my decisions. And I think the true transformative decision making happens when you're inspired by the data to do the right thing, not when you hope to be driven by it. And so that's the name of the book, Data Inspired. You can get it wherever books are sold. There's a website called datainspired.org and I'm a lot on LinkedIn, so feel free to follow. Feel free to not just follow, but also connect and reach out. I'm also always happy to learn from other data enthusiasts around the world. That's what I do and that's what I love.
- [34:01]
C
I love the data inspired concept because it still tells us that we are in charge, which I totally believe that 100%. So, Sebastian, thanks for coming on the show so much today.
- [34:15]
B
This has been an absolute pleasure. Thank you so much.
- [34:19]
A
Thank you, Sebastian, for the insightful interview. This is what I take away from this interview. The first lesson is that data only changes organizations when leadership changes the rules around how decisions are made. The second is that incentives matter as much as analytics. If people are rewarded for safety, they will avoid challenge, and if they are rewarded for visibility, they will hide uncertainty. The third is that AI does not fix weak decision making. It accelerates whatever culture already exists. For enterprise leaders, that means asking hard questions. Are we using data to explore better options or to protect current ones? Do our meetings reward curiosity or compliance? Where should humans stay accountable even when automation is available? The broader trend here is that technology is moving faster than organizational maturity. That gap will define the winners over the next several years. The organizations that succeed will not be the ones with the most models or dashboards. They will be the ones that know how to think, decide, and adapt with discipline. Data makes decision making better, but only leadership makes it meaningful.
- [35:36]
C
Thanks for listening to Embracing Digital Transformation. If you enjoyed today's conversation, give us five stars on your favorite part, podcasting app, or on YouTube. It really helps others discover the show. If you want to go deeper, join our exclusive community@patreon.com embracingdigital where we share bonus content and you can always connect with other change makers like yourself. You can always find more resources@embracingdigital.org until next time, keep embracing the digital transformation.